by llogiq

llogiq / bytecount

Counting occurrences of a given byte or UTF-8 characters in a slice of memory – fast

138 Stars 17 Forks Last release: Not found Other 106 Commits 0 Releases

Available items

No Items, yet!

The developer of this repository has not created any items for sale yet. Need a bug fixed? Help with integration? A different license? Create a request here:


Counting bytes really fast

Build Status Windows build status Current Version License: Apache 2.0/MIT

This uses the "hyperscreamingcount" algorithm by Joshua Landau to count bytes faster than anything else. The newlinebench repository has further benchmarks for old versions of this repository.

To use bytecount in your crate, if you have cargo-edit, just type

cargo add bytecount
in a terminal with the crate root as the current path. Otherwise you can manually edit your
to add
bytecount = 0.5.1
to your

In your crate root (
, depending on if you are writing a library or application), add
extern crate bytecount;
. Now you can simply use
as follows:
extern crate bytecount;

fn main() { let mytext = "some potentially large text, perhaps read from disk?"; let spaces = bytecount::count(mytext.as_bytes(), b' '); .. }

bytecount supports two features to make use of modern CPU's features to speed up counting considerably. To allow your users to use them, add the following to your

runtime-dispatch-simd = ["bytecount/runtime-dispatch-simd"]
generic-simd = ["bytecount/generic-simd"]

The first,

, enables detection of SIMD capabilities at runtime, which allows using the SSE2 and AVX2 codepaths, but cannot be used with

Your users can then compile with runtime dispatch using:

cargo build --release --features runtime-dispatch-simd

The second,

, uses
to provide a fast architecture-agnostic SIMD codepath, but requires running on nightly.

Your users can compile with this codepath using:

cargo build --release --features generic-simd

Building for a more specific architecture will also improve performance. You can do this with

RUSTFLAGS="-C target-cpu=native" cargo build --release

The scalar algorithm is explained in depth here.

Note: Versions until 0.4.0 worked with Rust as of 1.20.0. Version 0.5.0 until 0.6.0 requires Rust 1.26 or later, and at least 1.27.2 to use SIMD. Versions from 0.6.0 require Rust 1.32.0 or later.


Licensed under either of at your discretion:

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.